應用於醫學影像處理之FPGA運算平台建置

Building a FPGA Computing Platform for Medical Image Processing

曾煥宗(Huan-Tsung Tseng)

指導教授 ： 蔡育秀

中原大學/工學院/醫學工程研究所/碩士(2009年)

https://doi.org/10.6840/cycu200900432

摘要

近年來，醫學影像為了要在固定感應面積內用更多的像素去表示生理組織，造成醫學影像的資料量越來越龐大，而非一般運算系統可以負擔；故本研究將利用FPGA設計多個運算節點平行處理，實現高速醫學影像處理系統。
為了讓使用者可以更容易開發醫學影像處理演算法，本系統設計多種醫學影像處理IP(Intellectual Property)並驗證其功能，並利用這些IP建立醫學影像處理資料庫，讓使用者可利用資料庫中多種影像處理IP實現演算法的最佳化，同時縮短系統開發時程、減低系統複雜度、提高系統開發的成功率。最後將利用FPGA內部微處理器實現嵌入式系統，讓使用者將影像處理演算法整合至FPGA，實現嵌入式醫學影像處理系統。
本系統將實現數位乳房影像處理以驗證此系統的可行性；使用資料庫中多種影像處理IP逐步實現乳房影像處理演算法，並利用硬體模擬(hardware co-simulation)驗證演算法功能正確，最後測試該演算法於通用型處理器及FPGA之運算效能。

關鍵字

FPGA ； 嵌入式系統 ； 乳房影像處理 ； 智慧財產權

並列摘要

In recent years, medical image systems utilize high resolution pixels to express body tissue structures. The huge amount image processing requirement became a burden to most general purpose computer system. In this study, an FPGA-based, scalable number crunching system is built to realize high-speed medical image processing applications.

To simplify the implementation of medical image processing applications, the system provides varieties image processing intellectual properties (IP). These IP libraries provide optimized processing algorithms and shorten the timeline to construct an image applications. The embedded microcontroller in the FPGA is also used to coordinate the integrated image processing paradigms and to realize the embedded high speed medical image processing system.

To verify and validate the system, a mammogram processing application is implemented. The algorithms, stepwise development and hardware co-simulation procedures are comprehensively described. The performance comparisons by the general purpose CPU system and the FPGA based system are also presented

並列關鍵字

FPGA ； IP ； mammogram image processing ； embedded system ； Intellectual Property

參考文獻

1. H.P. Chan, L.T. Niklason, D.M. Ikeda, K.L. Lam, and D.D. Adler, 「Digitization requirements in mammography: Effects on computer-aided detection of microcalcifications,」 Med. Phys. 21, 1994, pp. 1203-1211.

2. J. Ge, B. Sahiner, L.M. Hadjiiski, H.-P. Chan, J. Wei, M.A. Helvie, and C. Zhou, 「Computer aided detection of clusters of microcalcifications on full field digital mammograms,」 Med. Phys.33, August 2006, pp. 2975-2988.

7. K. Herrmann, M. Seifert, K. Gaedke, H. Jeschke, and P. Pirsch, 「Architecture and VLSI Implementation of a RISC Core for a Monolithic Video Signal Processor,」 VLSI Signal Processing VII, January, 1994, pp. 368-377.

8. F. Zhou, J. Yang, M. Ling, and L. Shi, 「Design of Fixed-point Multimedia Hardware Accelerator in Specific SoC Based on Statistical Analysis,」 Proceedings of the Communications, Circuits and Systems, May 2005, pp. 1338-1342.

11. J. Detrey, and F.d. Dinechin, 「A parameterizable floating-point logarithm operator for FPGAs,」 Signals, Systems and Computers, 2005, pp. 1186-1190.